The population of women who survive breast cancer is rising rapidly. Despite the success of initial treatments, many patients constantly battle fears of disease recurrence and early death. However, the effect of non-clinical factors, particularly genetic factors, on breast cancer outcomes is largely unknown. The proposed study will use the existing resources of two well-established cohort studies of 3593 breast cancer patients to comprehensively evaluate the following hypotheses: 1) Breast cancer survival may be associated with genetic polymorphisms in genes encoding angiogenic factors and matrix metalloproteinases, both of which are essential for tumor growth and metastasis. 2) Infiltrating inflammatory cells, particularly tumor- associated macrophages, can produce a large variety of promalignant cytokines and growth factors. Genetic polymorphisms in inflammatory chemokine and cytokine genes may be related to breast cancer survival. 3) Transforming growth factor-B promotes the growth and progression of breast cancer, and genetic polymorphisms in TGF-R pathway genes may be related to breast cancer survival. 4) The cyclooxygenase-2 (COX2) gene is up-regulated in a large proportion of mammary tumors, and this enzyme initiates the biosynthesis of various prostaglandins with diverse, and sometimes opposing, effects on tumorigenesis. Genetic polymorphisms of prostaglandin-pathway genes may be associated with breast cancer survival. A two-phase study design will be applied. In Phase I, all functional variants plus haplotype-tagging SNPs will be genotyped among 1193 breast cancer patients who have been followed for an average of 7.1 years. All promising associations identified in Phase I will be evaluated in Phase II in an on-going cohort study of 2400 cancer patients (being followed for 5 years). The large sample size and two-phase study design will balance both Type I and Type II errors and provide credible results towards improving the understanding of associations between genetic factors and breast cancer outcomes. Identifying factors that predict risk of relapse and rates of mortality will not only affect the expanding cancer survivor population by providing evidence-based information, but will also positively influence the medical care system and economy at large by making treatment more effective and cost-efficient. The proposed study, built on successfully implemented cohort studies, will be extremely timely and cost-efficient.